EPA guidance on the early intervention in clinical high-risk states of psychoses

Transcription

1 EPA guidance on the early intervention in clinical high-risk states of psychoses Stefanie J. Schmidt 1, Frauke Schultze-Lutter 1, Benno G. Schimmelmann 1, Nadja P. Maric 3, Raimo R.K. Salokangas 4, Anita Riecher-Rössler 5, Mark van der Gaag 6, Anna Meneghelli 7, Merete Nordentoft 8, Max Marshall 9, Anthony Morrison 10, Joachim Klosterkötter 2, Stephan Ruhrmann 2 1 University Hospital of Child and Adolescent Psychiatry and Psychotherapy, University of Bern, Bern, Switzerland 2 Department of Psychiatry and Psychotherapy, University of Cologne, Cologne, Germany 3 School of Medicine, University of Belgrade and Clinic of Psychiatry, Clinical Center of Serbia, Belgrade, Serbia 4 Turku, Finland 5 Basel, Switzerland 6 Amsterdam, The Netherlands 7 Milano, Italy 8 Copenhagen, Denmark 9 Manchester, UK 10 Manchester, UK 1

2 Abstract The aim of this guidance paper of the European Psychiatric Association (EPA) is to provide evidence-based recommendations on the early intervention in clinical high-risk (CHR) states of psychoses as assessed according to the EPA guidance on early detection. These recommendations were derived from the current empirical evidence and a meta-analysis on the efficacy of psychological and pharmacological interventions in CHR samples. Studies had to investigate conversion rate and/or functioning as treatment outcome in CHR patients defined by the ultra-high risk and/or basic symptom criteria. Besides analyses on treatment effects on conversion rate and functional outcome, age and intervention approach were examined as potential moderators. Based on data of 15 studies (n=1394), early intervention approaches generally produced significantly reduced conversion rates at 6- to 48-month follow-ups compared to the control conditions. However, they failed to achieve significantly greater functional improvements because both early intervention and control conditions produced similar positive effects. With regard to the intervention approach, both psychological and pharmacological interventions produced significant effects on conversion rates but not on functional outcome relative to the control conditions. Early intervention in youth samples was generally less effective than in predominantly adult samples. Six evidence-based recommendations for an early intervention in CHR samples could already be formulated although more studies to investigate the specificity of treatment effects and potential age effects in order to tailor interventions to the individual treatment needs and risk status are clearly needed. Key words: prevention, early intervention in Europe, meta-analysis, risk, psychosis, adolescents, youth, cognitive-behavioral therapy, antipsychotics, neuro-protective 2

3 1. Introduction Each year 38.2% of the population of the European Union, i.e., million persons, suffer from any mental disorder (Wittchen 2011). This is associated with huge societal and individual burden (Gustavsson 2011; Whiteford 2013). Prevention has therefore become an integral part of European and international health care policies in order to reduce the prevalence and burden of mental disorders across the lifespan (EC, 2005, 2008; Campion 2012; Haro 2014) Functional disability in psychotic disorders Schizophrenia is among the seven leading causes of years lost to disability (YLDs) in adults in Europe (Wittchen 2011). This is mainly due to the fact that functional recovery rates have not changed substantially over the past 25 years, despite advances in pharmacological and psychological treatments (Jääskeläinen 2013, Shivashankar 2013). As a consequence, prevention of schizophrenia and psychotic disorders in general has attracted special interest (Solis, 2014). Functional impairments are already present before (Addington 2011) and often worsen until the onset of psychosis (Häfner 1999). Furthermore, they are one of the main predictors of poor clinical outcome including conversion to psychosis (Granö 2014). This emphasizes the need to intervene as early as possible to avoid or at least diminish these burdens and thereby to prevent transition to manifest psychosis Prevention in clinical high risk states of psychoses In psychosis research, an indicated prevention approach has been adopted that targets helpseeking persons who experience early signs of the emerging psychosis but do not meet diagnostic criteria with the ultimate goal to prevent the incidence of the disorder (McGlashan & Johannesson, 1999; McGorry 1998, 2002). Thus, indicated prevention faces two 3

4 challenges: (1) the accurate identification of the target population and (2) their effective treatment. For the purpose of early detection, two complementary sets of clinical high risk (CHR) criteria have been developed: the basic symptom (Huber, 1966; Schultze-Lutter, 2009) and the ultra-high risk (UHR) approach (Miller 2003; Yung 2005). Their evidence is systematically reviewed by and recommendations for their use are given in the accompanying European guidance on early detection (see Schultze-Lutter this issue). Notably, fulfilling these criteria only indicates an increased risk for developing psychosis which is always associated with an error probability resulting in false-positive predictions (Ruhrmann 2010). This has fueled ongoing debates about the risk of negative effects associated with the identification and treatment of CHR states of psychoses (e.g., stigmatization, financial loss) (e.g., Fusar-Poli 2014; Nelson, 2014; ethics-chapter) Need for treatment in clinical high risk states of psychoses In addition to the markedly increased risk for developing psychosis (Schultze-Lutter this issue), the most important argument in favor of an intervention in CHR patients is the reported distress and stigmatization caused by their mental problems already at the time of referral to an early detection and intervention service (Kline 2014; Rüsch 2014a,b; Stowkowy et al., 2014). This is reflected by their even higher levels of hopelessness, depressiveness, anxiety, and poor quality of life in comparison to other help-seeking patients and patients with firstepisode psychosis (Bechdolf 2005; Lee 2011; Pruessner 2011; Granö 2014a,b). Accordingly, they often fulfill the diagnostic criteria for another mental disorder, in particular for depression, anxiety, and substance abuse or dependence, which require treatment (Woods 2009; Hui 2013; Fusar-Poli 2014). Furthermore, CHR patients exhibit poor coping skills, low self-efficacy, and excessive external attributions that resemble profiles of depressive patients and might unfavorably interact with their frequent depressive mood (Schmidt 2014). 4

5 In addition, CHR patients demonstrate abnormalities in neuro- and social cognition with performances that are usually intermediate between those of healthy controls and schizophrenia patients (Fusar-Poli 2013; Giuliano 2013; Thompson 2013; Bora 2014; Brent 2014), and are associated with markedly impaired functional outcome and negative symptoms (Addington 2009; Lin 2011; Carrion 2013; Olvet 2013; Granö 2014; Salokangas 2014). Consistent with the conceptualization of psychoses as neurodevelopmental disorders, these neuro- and social-cognitive abnormalities are also accompanied and possibly reflected by various abnormalities in functional and structural imaging (Smieskova 2013; Bois 2014), in electrophysiological measures (Higuchi 2013; Kayser 2014) and neurochemistry (Leweke 2012; Egerton 2014; Gran 2014). In summary, CHR patients are independent of any potential risk to develop psychosis in the future certainly in need for treatment (Ruhrmann et al. 2010) Requirements for early intervention approaches In accordance with this obvious need for treatment, an increasing number of interventions have been evaluated in CHR samples in recent years (Ruhrmann 2012; Okuzawa 2014; Stafford 2014). With the primary goal to prevent conversion to psychosis, they have mainly built upon well-established interventions for adult schizophrenia patients and used conversion to psychosis as their primary outcome (McGorry 2002; Morrison 2004). Other more recently developed interventions have taken into account that CHR patients not only suffer from risk symptoms but also from several other mental problems and have targeted a broader array of outcomes in various settings with various intervention techniques (e.g., intensive case management, multi-family psychoeducation) (Granö 2014a,b; Marvin 2014). Yet most new generation intervention studies have an uncontrolled single-group design, therefore lack methodological rigor, and were not included in recent meta-analyses on the efficacy of 5

6 randomized controlled trials (RCTs) in CHR states (Cella & Preti, 2010; Marshall & Rathbone, 2011; Fusar-Poli 2013; van der Gaag, 2013; Stafford 2013; Hutton & Taylor, 2014). Current guidelines have not considered these new generation interventions (DGPPN, IEPA; NICE). Moreover, no sufficient evidence-based differential indication for the available interventions exists, which would require the examination of the treatment approach as a potential moderator variable in meta-analyses and/or in direct comparison in large RCTs with multiple treatment arms. In this context, age should also be studied as a potential moderator because CHR samples commonly include adolescents and young adults who differ in their social, emotional, and cognitive developmental state Aims The main aim of this guidance paper on the early intervention in CHR states was therefore to evaluate the efficacy of intervention approaches in CHR patients that focus on both prevention of conversion to psychosis and improvement in functional outcome with special consideration of the potential moderating effects of age and intervention type. This served the ultimate goal to derive evidence-based recommendations on early intervention in CHR states assessed according to the recommendations provided by the EPA guidance on their early detection (see [Schultze-Lutter et al.], this issue). 2. Methods 2.1. Study selection Literature search We conducted a systematic literature search in June 2014 in PubMed (no time limit), PsycInfo (no time limit), Scopus (no time limit) that covers all journals included in Embase, and in the Cochrane Collaboration Controlled Trials Register using the following search terms and syntax: ((prevention) OR (early intervention) OR (treatment) OR (therapy)) AND ((risk) OR (prodrome) 6

7 AND ((psychosis) OR (schizophrenia)). Furthermore, we inspected reference lists of all identified reviews and meta-analyses. Extracted outcome Selection criteria TR (PANSS) TR (SIPS), FO (GAF & SFS) TR (CAARMS or reports from family We included all studies in our meta-analysis that (i) evaluated a psychological and/or pharmacological intervention including interventions with nutritional supplements and other substances, (ii) included a majority (>50%) of CHR patients as defined by the UHR and/or the basic symptom criteria, (iii) reported conversion rates and/or functional outcome as intervention outcome, and (iv) were published in English. Exclusion criteria were: (i) studies with samples that were also part of other studies included in our analyses with a larger sample size and/or a longer follow-up period, (ii) studies that were only published as an abstract or trial protocol, (iii) case-reports, and (iv) studies that used an observational naturalistic design without any specific intervention Selection process and quality assessment Tab. 1 Characteristics of studies included in the meta-analysis Psychological interventions (PSY) Follow-up (months after baseline); Dropout-rate (Posttherapy) Control group Design Intervention Sample characteristics Sample size In- & exclusion criteria Study Coun -try 6, 12, 36; EG / CG: 30% / 30% Monitoring; monthly CBT + monitoring; 26 sessions, 6 months - Age (yrs.): EG: 20.6±4.9 / CG: 21.5±5.2 (Age group: ADULT) - Gender: (male): EG: 60% / CG: 83% - Co-morbidities: not reported 60 EG: 37 CG: 23 UK RCT Inclusion criteria: - Risk for psychosis (PANSS) - < 16 years, > 36 years - Current or past receipt of antipsychotic medication Morrison 2004 [x], 2007 GE 1-7 6, 12, 18; EG / CG: 30% / 33%, Supportive Therapy: coping with current problems, psychoeducation; 20 sessions, 6 months CBT; max. 20 sessions (mean=12±6.2, range=1-26), 6 months 20.8±4.5 / CG: 21.1±3.7 (Age group: ADULT) - Gender (male): EG: 67% / CG 75% - Co-morbidities (EG / CG): mood disorders: 26% / 26%, alcohol abuse: 18% / 18%, cannabis abuse: 10% / 10% 51 EG: 27 CG: 24 CAN RCT Inclusion criteria - 14 to 30 years - Risk for psychosis (SIPS) Exclusion criteria - Lifetime or current axis-i psychotic disorder - Prior treatment with an antipsychotic - IQ < 70 - Past/current central nervous system disorder Addington 2011 Marshall 2012 GE 1-6, 12, 18, 24; EG / CG: 33% / 31% Monitoring; monthly CBT + Monitoring; Max ±4.2 / CG: 20.8±4.5 (Age 288 EG: 144 CG: UK RCT Inclusion criteria: - At-risk for psychosis (CAARMS) Morrison 2012,

8 Morrison 2011, 2013 GE 1 + Bechdolf 2012 Bechdolf 2007 GE 1 - GER years - Help-seeking - Current or previous receipt of antipsychotic drugs - Moderate to severe learning disability - Organic impairments - Insufficient English RCT Inclusion criteria: - At least one of 10 thought or perceptional basic symptoms (ERIraos) and/or - Reduction in the GAF Score (DSM-IV) of at least 30 points within the past year and at least one of these risk factors: first-degree relative with schizophrenia /schizophrenia spectrum disorder or pre- /perinatal complications; - APS or BLIPS - Present or past diagnosis of a psychotic disorder, bipolar disorder, organic brain 144 group: ADULT) - Gender: (male): EG: 62% / CG: 63% - Co-morbidities (total sample, >5%): Depressive disorder: 34%, dysthymic disorder: 7%, panic disorder with agoraphobia: 6%, panic disorder without agoraphobia: 11%, social phobia: 11%, specific phobia: 11%, generalized anxiety disorder: 9%, obsessive compulsive disorder: 8% 128 EG: 63 CG: ±5.4 / CG: 26.8±6.2 (Age group: ADULT) - Gender: (male): 62% / 65% - Co-morbidities: not reported sessions, 6 months; plus up to 4 boostersessions in the following 6 months; Integrated treatment: : Individual CBT, multifamily psychoeducation (group), social skills training (group), cognitive remediation 25 sessions, 12 months Supportive counselling: coping with current problems, basic psychoeducation; 30 sessions, 12 months 6, 12, 18, 24; EG / CG: 19%, / 12% doctors), FO (GAF) TR (DSM-IV), FO (SAS II)

9 Van der Gaag et al., 2012 Rietdijk et al., 2010 GE 1 ++ McGorry 2013 Yung et al GE 1 + NL AUS disorder, substance dependence - Mental retardation - Previous treatment with antipsychotics - Acute suicidality - < 17 years, > 35 years RCT Inclusion criteria - 14 to 35 years - At-risk for psychosis (CAARMS 2006) - SOFAS score 50 and/or a reduction by 30% for at least 1 month in the past year - Current or previous use of antipsychotic medication with 15 mg cumulative haloperidol equivalent - Severe learning impairment - Problems due to an organic condition - Insufficient competence in Dutch; - History of psychosis RCT Inclusion criteria: - Age years - Melbourne metropolitan area - Risk for psychosis (CAARMS 2006) - History of psychotic or manic episode - Medical condition that accounts for symptoms - Neurologic, biochemical or 201 EG: 98 CG: EG: 44 CG: ±5.6 / CG: 22.6±5.5 (Age group: ADULT) - Gender (male): EG: 50% / CG: 49% - Co-morbidities: (total sample, > 5%): Anxiety disorders: 27%, depression: 26%, personality disorders: 8%, ADHD: 7%, addiction problems: 6%; 18.0±2.7 / CG: 18.8±3.7 (Age group: YOUTH) - Gender (male): EG: 49% / CG: 47% - Co-morbidities: not reported CBT + Treatment As Usual (TAU); max. 26 sessions, weekly; 6 months; additional evidencebased treatment for axis-i and axis-ii disorders; EG : CBT + Placebo; 12 months; CBT: Weekly to monthly basis; min./session with the number of sessions not determined Treatment As Usual (TAU); additional evidencebased treatment for axis-i and axis-ii disorders; Supportive therapy + placebo; 12 months; 6, 12, 18; EG / CG: 15% / 12% 6, 12; EG: 34% / CG: 32% TR (CAARMS 2006), FO (SOFAS) TR (CAARMS 2006), FO GAF)

10 Miklowitz 2014 O Brien et al., 2014 GE 1 + O Brien et al., 2007 GE 2 - USA USA hematologic abnormalities - Serious co-existing illnesses - Lifetime antipsychotic dose of 15mg or more of haloperidol - Any previous or current use of mood-stabilizing medication - History of severe drug allergy - IQ<70 - Pregnancy or lactation - Insufficient English RCT Inclusion criteria: - 12 and 25 years - Speaks and writes English - At-risk for psychosis (SIPS) - Current schizophrenia or schizoaffective disorders - Developmental disorders - Substance use disorders - Neurological disorders No CG Inclusion criteria: years - At-risk for psychosis (SIPS) - DSM-IV diagnosis of a schizophrenia spectrum disorder - IQ<70 - Current drug or alcohol dependence 129 EG: 66 CG: ±4.2 / CG: 17.4±3.9 (Age group: YOUTH) - Gender (male): EG: 59 %/ CG: 56% - Co-morbidities (EG / CG, >10%): depressive disorder (40%/ 29%), anxiety disorders (42%/ 50%), ADD (21%/ 18%), learning disorders (11%/ 7%) 16 - Age (yrs.): 15.7 (range: ) (Age group: YOUTH) - Gender: (male): 50% - Comorbidities (>10%): Mood disorders: 63%, depressive in advance; Family focused treatment; 18 sessions à 60 minutes, 6 months Psychoeducational multi-family group; Weekly sessions à 90 minutes, 9 months Enhanced care; 3 weekly psychoeducational sessions, 1 month 6; EG: 17% / CG: 25% - 9; 45% declined or dropped out TR (SIPS), FO (GAF) FO (GAF)

11 Hooker et al., 2014 GE 2 - USA No CG - Current neurological disorder disorder: 31%, depressive disorder NOS: 19%, anxiety disorder NOS: 31%, generalized anxiety disorder: 19%, ADHD: 13%, eating disorder NOS: 19% / 0% Inclusion criteria: years - At-risk for psychosis (SIPS) - Major medical / neurological illness - Non-fluent English - MR-contraindication - IQ<70 28 EG: 14 CG: 14 - Age (yrs.): 21.9±4.2 (Age group: ADULT) - Gender (male): 50% Pharmacological studies - with antipsychotics (MED) McGlasha n 2006 McGlasha n 2003 Woods et al., 2003 GE 1 + Woods et al., 2007 USA RCT Inclusion criteria: years - Help-seeking - At-risk for psychosis (SIPS) - Past or current psychotic disorder - Treatable psychiatric disorder that could account for the prodromal symptoms - Suicidal or homicidal - Prodromal symptoms due to drug/alcohol use USA No CG Inclusion criteria: years - Treatment seeking outpatients 60 EG:31 CG: ±5.5 / CG: 17.2±4.0 (Age group: YOUTH) - Gender (male): EG: 68% / CG: 62% - Co-morbidities: Current substance abuse or dependence (EG: 13% / CG: 4%) 15 - Age (yrs.): 17.1±5.5 (Age group: YOUTH) - Gender (male): 53% CRT; neuroand socialcognitive computerize d exercises; 1h each day, 5 days/week, 8 weeks; Olanzapine; 5-15 mg/d, 12 months; Additional individual and family psychosocial treatment, varied across sites; Aripiprazole ; Initial doses were 1st - 3; 18% Placebo; 12 months; Additional individual and family psychosocia l treatment, varied across sites; - 2; 13% 2, 12, 24; EG: 55% / CG: 35% FO (Global Functioning Role and Social scales) TR (SIPS), FO (GAF) FO (GAF, SFS)

12 GE 2 - Tsujino et al., 2013 GE 2 - JPN No CG - Met diagnostic criteria for a possible prodromal syndrome (SIPS) - Past or current DSM-IV criteria for any lifetime psychotic disorder - Psychiatric disorder which could account for the symptoms - Symptoms primarily as sequelae to drug or alcohol use - Alcohol or drug misuse or dependence in the past 3 months - Use of antipsychotic medication in the previous 3 months - Change in dosage of any antidepressant within 6 weeks, stimulant medication within 4 weeks or mood stabiliser within 4 weeks Inclusion criteria: years - Help-seeking outpatients - At-risk for psychosis (SIPS) - Previous diagnosis of any psychotic disorder (DSM-IV) - Symptoms fully accounted for by an Axis 1 disorder or sequelae of drug/alcohol use - Abuse of alcohol or drugs - Antipsychotic medication use - Co-morbidities: not reported 11 - Age (yrs.): 26.7±6.5 (Age group: ADULT) - Gender (male): 46% - Co-morbidities: not reported Week: 5 mg/d, 2nd Week: 10 mg/d 3rd week: 15 mg/d, 4th Week: 20mg/d and if needed to 30mg/d; 6 weeks Perospirone ; dosing according to a flexible schedule; psychosocial therapy available; 26 weeks - 6; 25% FO (GAF)

13 Pharmacological studies - combined with psychological interventions (MED) McGorry et al Phillips 2007 GE 1 - Ruhrmann 2007 GE 1 - AUS GER RCT Inclusion criteria: years - Live in the Melbourne metropolitan area - Risk for psychosis (CAARMS) - Previous psychotic or manic episode - Previous treatment with an antipsychotic or mood stabilizing agent - Substance-induced psychotic disorder - IQ<70 - Inadequate command of English RCT Inclusion criteria: - Older than 18 years - Risk for psychosis (ERIraos) - Lifetime DSM IV diagnosis of schizophrenia spectrum disorder, brief psychotic episode (>1 week), delirium, dementia, amnestic and other cognitive disorders - Mental retardation - Mental disorders due to a general medical condition or psychotropic substances - Abuse of alcohol or drugs within the past 3 months or the past 4 weeks for cannabis - Any lifetime continuous treatment with high-potency 59 EG: 31 CG: EG: 65 CG: 59 20±4/ CG: 20±3 (Age group: ADULT) - Gender (male): EG: 65% / CG: 50% 25.1±6.6 / CG: 26.1±6.1 (Age group: ADULT) - Gender: (male): 48% / 60% - Co-morbidities: not reported Needsfocused intervention + Amisulpride; 12 weeks; mg/d, with increments of 50 mg at first step and 100 mg at further steps; dosage was increased as long as APS and BLIPS were Risperidone (1-2 mg/d) + CBT + Needsbased intervention (NBI); 6 months; NBI ongoing; Needsbased interventio n (NBI); 6 months; NBI ongoing; Needsfocused interventio n; 12 weeks 6, 12, 36-48; months; Drop-out rate: not reported, 41% nonadherent to Risperidone; 3; EG: 29% / CG: 49% FO (GAF) FO (GAF)

14 McGorry 2013 Yung et al GE 1 + AUS antipsychotics (> 1 week) or antipsychotics during 6 months prior to the study - Any contraindication for amisulpride - Women of childbearing risk not using contraception RCT Inclusion criteria: - Age years - Melbourne metropolitan area - Risk for psychosis (CAARMS, 2005) - History of psychotic or manic episode - Medical condition that accounts for symptoms - Neurologic, biochemical or hematologic abnormalities - Serious co-existing illnesses - Lifetime antipsychotic dose of 15mg or more of haloperidol - Any previous or current use of mood-stabilizing medication - History of severe drug allergy - IQ<70 - Pregnancy or lactation - Insufficient English 71 EG: 43 CG: ±3.0) / CG: 18.8±3.7 Age group: YOUTH) - Gender (male): EG: 45% / CG: 47% - Co-morbidities: not reported Pharmacological studies - with nutritional supplements (MED) Amminger 2010 Mossaheb 2013 AUT RCT Inclusion criteria: - At-risk for psychosis (PANSS) Exclusion Criteria: - History of previous psychotic disorder or manic episode - Substance-induced psychotic 81 EG: 41 CG: ±2.4 / CG: 16.0±1.7 (Age group: YOUTH) - Gender (male): EG: 34% / CG: 33% present; EG: CBT + Risperidone (0.5-2 mg/d); 12 months; CBT: Weekly to monthly basis; min./session ; number of sessions not predetermin ed; 1.2 g/d ω-3 PUFAS; 12 weeks; 9 additional sessions of psychologic Supportive therapy + placebo; 12 months; Placebo (coconut oil); 12 weeks; psychologic al and 6, 12; EG 37% / CG: 32% 12; EG: 7% / 5%; TR (CAARMS 2006), FO GAF) TR (PANSS), FO (GAF)

15 GE 1 + disorder - Acute suicidal or aggressive behavior - Current DSM-IV diagnosis of substance dependence (except cannabis dependence) - Neurological disorders - IQ<70 - Structural brain changes apparent on magnetic resonance imaging - Previous treatment with an antipsychotic or moodstabilizing agent (>1 week) - ω-3 supplements within 8 weeks of being included in the trial - Laboratory values more than 10% outside the normal range for transaminases, thyroid hormones, C-reactive protein, or bleeding parameters - Another severe intercurrent illness - Co-morbidities: not reported al and psychosocial intervention s; psychosocia l interventio ns;

16 Table 2 Within-group effect sizes at different follow-ups for improvements in functional outcome Fo St Ad Ad Am Be Ho Ho M M M M M M O` Ru Ts Va W W Po (g He Be

17 p=0.83 p=0.29 p=0.91 Note. * < 0.05, ** < 0.01, ***< 0.001; Effect sizes are presented in a way that positive values indicate an improvement in functional outcome. Abbreviations: CG, control group; EG, experimental group; g w : standardized mean difference for pre-post improvements in the respective group;

18 Tab. 1 Characteristics of studies included in the meta-analysis Psychological interventions (PSY) Study Morrison 2004 [x], 2007 GE 1 - Addington 2011 Marshall 2012 GE 1 - Morrison 2012, Coun -try UK In- & exclusion criteria RCT Inclusion criteria: - Risk for psychosis (PANSS) - < 16 years, > 36 years - Current or past receipt of antipsychotic medication CAN RCT Inclusion criteria - 14 to 30 years - Risk for psychosis (SIPS) Exclusion criteria - Lifetime or current axis-i psychotic disorder - Prior treatment with an antipsychotic - IQ < 70 - Past/current central nervous system disorder UK RCT Inclusion criteria: - At-risk for psychosis (CAARMS) Sample size 60 EG: 37 CG: EG: 27 CG: EG: 144 CG: Sample characteristics - Age (yrs.): EG: 20.6±4.9 / CG: 21.5±5.2 (Age group: ADULT) - Gender: (male): EG: 60% / CG: 83% - Co-morbidities: not reported 20.8±4.5 / CG: 21.1±3.7 (Age group: ADULT) - Gender (male): EG: 67% / CG 75% - Co-morbidities (EG / CG): mood disorders: 26% / 26%, alcohol abuse: 18% / 18%, cannabis abuse: 10% / 10% 20.7±4.2 / CG: 20.8±4.5 (Age Design Intervention CBT + monitoring; 26 sessions, 6 months CBT; max. 20 sessions (mean=12±6.2, range=1-26), 6 months CBT + Monitoring; Max. 26 Control group Monitoring; monthly Supportive Therapy: coping with current problems, psychoeducation; 20 sessions, 6 months Monitoring; monthly Follow-up (months after baseline); Dropout-rate (Posttherapy) 6, 12, 36; EG / CG: 30% / 30% 6, 12, 18; EG / CG: 30% / 33%, 6, 12, 18, 24; EG / CG: 33% / 31% Extracted outcome TR (PANSS) TR (SIPS), FO (GAF & SFS) TR (CAARMS or reports from family

19 Morrison 2011, 2013 GE 1 + Bechdolf 2012 Bechdolf 2007 GE 1 - GER years - Help-seeking - Current or previous receipt of antipsychotic drugs - Moderate to severe learning disability - Organic impairments - Insufficient English RCT Inclusion criteria: - At least one of 10 thought or perceptional basic symptoms (ERIraos) and/or - Reduction in the GAF Score (DSM-IV) of at least 30 points within the past year and at least one of these risk factors: first-degree relative with schizophrenia /schizophrenia spectrum disorder or pre- /perinatal complications; - APS or BLIPS - Present or past diagnosis of a psychotic disorder, bipolar disorder, organic brain 144 group: ADULT) - Gender: (male): EG: 62% / CG: 63% - Co-morbidities (total sample, >5%): Depressive disorder: 34%, dysthymic disorder: 7%, panic disorder with agoraphobia: 6%, panic disorder without agoraphobia: 11%, social phobia: 11%, specific phobia: 11%, generalized anxiety disorder: 9%, obsessive compulsive disorder: 8% 128 EG: 63 CG: ±5.4 / CG: 26.8±6.2 (Age group: ADULT) - Gender: (male): 62% / 65% - Co-morbidities: not reported sessions, 6 months; plus up to 4 boostersessions in the following 6 months; Integrated treatment: : Individual CBT, multifamily psychoeducation (group), social skills training (group), cognitive remediation 25 sessions, 12 months Supportive counselling: coping with current problems, basic psychoeducation; 30 sessions, 12 months 6, 12, 18, 24; EG / CG: 19%, / 12% doctors), FO (GAF) TR (DSM-IV), FO (SAS II)

20 Van der Gaag et al., 2012 Rietdijk et al., 2010 GE 1 ++ McGorry 2013 Yung et al GE 1 + NL AUS disorder, substance dependence - Mental retardation - Previous treatment with antipsychotics - Acute suicidality - < 17 years, > 35 years RCT Inclusion criteria - 14 to 35 years - At-risk for psychosis (CAARMS 2006) - SOFAS score 50 and/or a reduction by 30% for at least 1 month in the past year - Current or previous use of antipsychotic medication with 15 mg cumulative haloperidol equivalent - Severe learning impairment - Problems due to an organic condition - Insufficient competence in Dutch; - History of psychosis RCT Inclusion criteria: - Age years - Melbourne metropolitan area - Risk for psychosis (CAARMS 2006) - History of psychotic or manic episode - Medical condition that accounts for symptoms - Neurologic, biochemical or 201 EG: 98 CG: EG: 44 CG: ±5.6 / CG: 22.6±5.5 (Age group: ADULT) - Gender (male): EG: 50% / CG: 49% - Co-morbidities: (total sample, > 5%): Anxiety disorders: 27%, depression: 26%, personality disorders: 8%, ADHD: 7%, addiction problems: 6%; 18.0±2.7 / CG: 18.8±3.7 (Age group: YOUTH) - Gender (male): EG: 49% / CG: 47% - Co-morbidities: not reported CBT + Treatment As Usual (TAU); max. 26 sessions, weekly; 6 months; additional evidencebased treatment for axis-i and axis-ii disorders; EG : CBT + Placebo; 12 months; CBT: Weekly to monthly basis; min./session with the number of sessions not determined Treatment As Usual (TAU); additional evidencebased treatment for axis-i and axis-ii disorders; Supportive therapy + placebo; 12 months; 6, 12, 18; EG / CG: 15% / 12% 6, 12; EG: 34% / CG: 32% TR (CAARMS 2006), FO (SOFAS) TR (CAARMS 2006), FO GAF)

21 Miklowitz 2014 O Brien et al., 2014 GE 1 + O Brien et al., 2007 GE 2 - USA USA hematologic abnormalities - Serious co-existing illnesses - Lifetime antipsychotic dose of 15mg or more of haloperidol - Any previous or current use of mood-stabilizing medication - History of severe drug allergy - IQ<70 - Pregnancy or lactation - Insufficient English RCT Inclusion criteria: - 12 and 25 years - Speaks and writes English - At-risk for psychosis (SIPS) - Current schizophrenia or schizoaffective disorders - Developmental disorders - Substance use disorders - Neurological disorders No CG Inclusion criteria: years - At-risk for psychosis (SIPS) - DSM-IV diagnosis of a schizophrenia spectrum disorder - IQ<70 - Current drug or alcohol dependence 129 EG: 66 CG: ±4.2 / CG: 17.4±3.9 (Age group: YOUTH) - Gender (male): EG: 59 %/ CG: 56% - Co-morbidities (EG / CG, >10%): depressive disorder (40%/ 29%), anxiety disorders (42%/ 50%), ADD (21%/ 18%), learning disorders (11%/ 7%) 16 - Age (yrs.): 15.7 (range: ) (Age group: YOUTH) - Gender: (male): 50% - Comorbidities (>10%): Mood disorders: 63%, depressive in advance; Family focused treatment; 18 sessions à 60 minutes, 6 months Psychoeducational multi-family group; Weekly sessions à 90 minutes, 9 months Enhanced care; 3 weekly psychoeducational sessions, 1 month 6; EG: 17% / CG: 25% - 9; 45% declined or dropped out TR (SIPS), FO (GAF) FO (GAF)

22 Hooker et al., 2014 GE 2 - USA No CG - Current neurological disorder disorder: 31%, depressive disorder NOS: 19%, anxiety disorder NOS: 31%, generalized anxiety disorder: 19%, ADHD: 13%, eating disorder NOS: 19% / 0% Inclusion criteria: years - At-risk for psychosis (SIPS) - Major medical / neurological illness - Non-fluent English - MR-contraindication - IQ<70 28 EG: 14 CG: 14 - Age (yrs.): 21.9±4.2 (Age group: ADULT) - Gender (male): 50% Pharmacological studies - with antipsychotics (MED) McGlasha n 2006 McGlasha n 2003 Woods et al., 2003 GE 1 + Woods et al., 2007 USA RCT Inclusion criteria: years - Help-seeking - At-risk for psychosis (SIPS) - Past or current psychotic disorder - Treatable psychiatric disorder that could account for the prodromal symptoms - Suicidal or homicidal - Prodromal symptoms due to drug/alcohol use USA No CG Inclusion criteria: years - Treatment seeking outpatients 60 EG:31 CG: ±5.5 / CG: 17.2±4.0 (Age group: YOUTH) - Gender (male): EG: 68% / CG: 62% - Co-morbidities: Current substance abuse or dependence (EG: 13% / CG: 4%) 15 - Age (yrs.): 17.1±5.5 (Age group: YOUTH) - Gender (male): 53% CRT; neuroand socialcognitive computerize d exercises; 1h each day, 5 days/week, 8 weeks; Olanzapine; 5-15 mg/d, 12 months; Additional individual and family psychosocial treatment, varied across sites; Aripiprazole ; Initial doses were 1st - 3; 18% Placebo; 12 months; Additional individual and family psychosocia l treatment, varied across sites; - 2; 13% 2, 12, 24; EG: 55% / CG: 35% FO (Global Functioning Role and Social scales) TR (SIPS), FO (GAF) FO (GAF, SFS)

23 GE 2 - Tsujino et al., 2013 GE 2 - JPN No CG - Met diagnostic criteria for a possible prodromal syndrome (SIPS) - Past or current DSM-IV criteria for any lifetime psychotic disorder - Psychiatric disorder which could account for the symptoms - Symptoms primarily as sequelae to drug or alcohol use - Alcohol or drug misuse or dependence in the past 3 months - Use of antipsychotic medication in the previous 3 months - Change in dosage of any antidepressant within 6 weeks, stimulant medication within 4 weeks or mood stabiliser within 4 weeks Inclusion criteria: years - Help-seeking outpatients - At-risk for psychosis (SIPS) - Previous diagnosis of any psychotic disorder (DSM-IV) - Symptoms fully accounted for by an Axis 1 disorder or sequelae of drug/alcohol use - Abuse of alcohol or drugs - Antipsychotic medication use - Co-morbidities: not reported 11 - Age (yrs.): 26.7±6.5 (Age group: ADULT) - Gender (male): 46% - Co-morbidities: not reported Week: 5 mg/d, 2nd Week: 10 mg/d 3rd week: 15 mg/d, 4th Week: 20mg/d and if needed to 30mg/d; 6 weeks Perospirone ; dosing according to a flexible schedule; psychosocial therapy available; 26 weeks - 6; 25% FO (GAF)

24 Pharmacological studies - combined with psychological interventions (MED) McGorry et al Phillips 2007 GE 1 - Ruhrmann 2007 GE 1 - AUS GER RCT Inclusion criteria: years - Live in the Melbourne metropolitan area - Risk for psychosis (CAARMS) - Previous psychotic or manic episode - Previous treatment with an antipsychotic or mood stabilizing agent - Substance-induced psychotic disorder - IQ<70 - Inadequate command of English RCT Inclusion criteria: - Older than 18 years - Risk for psychosis (ERIraos) - Lifetime DSM IV diagnosis of schizophrenia spectrum disorder, brief psychotic episode (>1 week), delirium, dementia, amnestic and other cognitive disorders - Mental retardation - Mental disorders due to a general medical condition or psychotropic substances - Abuse of alcohol or drugs within the past 3 months or the past 4 weeks for cannabis - Any lifetime continuous treatment with high-potency 59 EG: 31 CG: EG: 65 CG: 59 20±4/ CG: 20±3 (Age group: ADULT) - Gender (male): EG: 65% / CG: 50% 25.1±6.6 / CG: 26.1±6.1 (Age group: ADULT) - Gender: (male): 48% / 60% - Co-morbidities: not reported Needsfocused intervention + Amisulpride; 12 weeks; mg/d, with increments of 50 mg at first step and 100 mg at further steps; dosage was increased as long as APS and BLIPS were Risperidone (1-2 mg/d) + CBT + Needsbased intervention (NBI); 6 months; NBI ongoing; Needsbased interventio n (NBI); 6 months; NBI ongoing; Needsfocused interventio n; 12 weeks 6, 12, 36-48; months; Drop-out rate: not reported, 41% nonadherent to Risperidone; 3; EG: 29% / CG: 49% FO (GAF) FO (GAF)

25 McGorry 2013 Yung et al GE 1 + AUS antipsychotics (> 1 week) or antipsychotics during 6 months prior to the study - Any contraindication for amisulpride - Women of childbearing risk not using contraception RCT Inclusion criteria: - Age years - Melbourne metropolitan area - Risk for psychosis (CAARMS, 2005) - History of psychotic or manic episode - Medical condition that accounts for symptoms - Neurologic, biochemical or hematologic abnormalities - Serious co-existing illnesses - Lifetime antipsychotic dose of 15mg or more of haloperidol - Any previous or current use of mood-stabilizing medication - History of severe drug allergy - IQ<70 - Pregnancy or lactation - Insufficient English 71 EG: 43 CG: ±3.0) / CG: 18.8±3.7 Age group: YOUTH) - Gender (male): EG: 45% / CG: 47% - Co-morbidities: not reported Pharmacological studies - with nutritional supplements (MED) Amminger 2010 Mossaheb 2013 AUT RCT Inclusion criteria: - At-risk for psychosis (PANSS) Exclusion Criteria: - History of previous psychotic disorder or manic episode - Substance-induced psychotic 81 EG: 41 CG: ±2.4 / CG: 16.0±1.7 (Age group: YOUTH) - Gender (male): EG: 34% / CG: 33% present; EG: CBT + Risperidone (0.5-2 mg/d); 12 months; CBT: Weekly to monthly basis; min./session ; number of sessions not predetermin ed; 1.2 g/d ω-3 PUFAS; 12 weeks; 9 additional sessions of psychologic Supportive therapy + placebo; 12 months; Placebo (coconut oil); 12 weeks; psychologic al and 6, 12; EG 37% / CG: 32% 12; EG: 7% / 5%; TR (CAARMS 2006), FO GAF) TR (PANSS), FO (GAF)

26 GE 1 + disorder - Acute suicidal or aggressive behavior - Current DSM-IV diagnosis of substance dependence (except cannabis dependence) - Neurological disorders - IQ<70 - Structural brain changes apparent on magnetic resonance imaging - Previous treatment with an antipsychotic or moodstabilizing agent (>1 week) - ω-3 supplements within 8 weeks of being included in the trial - Laboratory values more than 10% outside the normal range for transaminases, thyroid hormones, C-reactive protein, or bleeding parameters - Another severe intercurrent illness - Co-morbidities: not reported al and psychosocial intervention s; psychosocia l interventio ns;

27 Table 2 Within-group effect sizes at different follow-ups for improvements in functional outcome Follow-up 2 to 6 months 9 to 12 months 18 months Study EG g w CG g w EG g w CG g w EG g w CG g w Addington x a Addington x b Amminger x Bechdolf x Hooker x a Hooker x b McGlashan x McGorry x McGorry x Risperidone McGorry x CBT Miklowitz x Morrison x O`Brien x 0.87 Ruhrmann x Tsujino x 0.97 Van der Gaag x Woods x a Woods x b 1.40 Pooled g w (g w, 95% CIs) 0.62*** (0.26,0.98) 0.68** (0.26,1.10) 0.84*** (0.41,1.26) 1.22*** (0.66,1.78) 0.69 (-0.01,1.39) 0.64* (0.12,1.17) Heterogeneity Q w(13)=101.65*** I 2 =87% Q w(19)=201.02*** I 2 =91% Q w(9)=67.27*** I 2 =87% Between-group differences Q b(1) = 0.04, p=0.83 Q b(1) = 1.12, p=0.29 Q b(1) = 0.01, p=0.91 Note. * < 0.05, ** < 0.01, ***< 0.001; Effect sizes are presented in a way that positive values indicate an improvement in functional outcome. Abbreviations: CG, control group; EG, experimental group; g w : standardized mean difference for pre-post improvements in the respective group;

28 Included Eligibility Screening Identification records identified through database searching Records after duplicates removed (n=8452) 3 records identified through other sources 77 records screened 16 records excluded 61 full-text articles assessed for eligibility 25 articles included in the review Studies included in the meta-analysis: articles excluded due to these reasons: Case-studies (n=4) Naturalistic design (n=4) Study protocol (n=10) Other outcome(s) (n=12) Mixed samples (n=3) Overlapping sample (n=3) Figure 1. Flow chart of the study selection process

29

30

31

32

33

34

35